COMPONENT ACTIVATION OF A HIGH CURRENT RADIOISOTOPE PRODUCTION MEDICAL CYCLOTRON

Bhaskar Mukherjee* and Joseph Khachan Department of Applied and Plasma Physics (A28), University of Sydney, NSW 2006, Australia

Abstract During routine operation of medical cyclotrons intense flux of fast neutrons and gamma rays are produced. The neutrons interact with the cyclotron parts, causing radio activation. The activated cyclotron parts pose consider-able radiological hazard during maintenance and waste disposal procedures. In this report the mechanism of cyclotron component activation and the long and short-term prediction thereof have been analysed and related important radiological safety aspects highlighted.

INTRODUCTION In July 1992 a 30 MeV high current (maximum proton current: 400 micro A) H--ion medical cyclotron Cyclone30 (IBA Belgium) has been installed at Royal Prince Alfred Hospital (RPAH), the prime teaching hospital of University of Sydney, jointly operated by ANSTO (Australian Nuclear Science and Technology Organisation). In March 1992 the cyclotron commenced its routine production of PET radio radiopharmaceuticals 11CO2, 13NH3, and 18FDG in aqueous form as well as the SPECT radioisotopes 201Tl, 67Ga and 123I.

During routine operation of medical cyclotrons high flux of fast neutrons are produced at cyclotron targets, where the energetic proton beams interact with the target material to produce the radioisotope of interest. The primary fast neutrons suffer multiple collisions with the vault wall, roof and floor surfaces resulting in their thermalisation via slowing down process [1]. These thermal neutrons fill up the cyclotron, in particular the target vault. Evidently, the cyclotron parts, “immersed” in the thermal neutron cloud suffer activation. In course of long term cyclotron operation, the accumulated activation of the cyclotron parts could cause considerable radiological risk, in particular during the maintenance, waste disposal and handling (i.e. cyclotron part replacement) processes. We have estimated the integrated thermal neutron flux at spots located at selected cyclotron part using cobalt activation pellets at different cyclotron operation conditions, scaled by the integrated target current in the beam-stopper. The pellets were assayed using a gamma spectrometer and the neutron fluences evaluated. Using the neutron fluence data and isotopic composition of the building martial we have calculated the activity of the radioactive species generated in the cyclotron part of interest and reconstructed the gamma dose rate at 1m. __________________________________________________________ *Present address:#DESY, Notkestrasse 85, D-22607 Hamburg, Germany Email: mukherjee@ieee.org

A practical method for the prediction of component activation and associated gamma dose rates of newly installed cyclotrons has been highlighted. The footprint of the 30 MeV medical cyclotron at RPAH/ANSTO is shown elsewhere [2].

MATERIALS AND METHODS Estimation of Neutron Fluence Tiny cobalt (59Co, isotopic abundance 100%) palettes (d = 8mm, t = 1mm, w = 447 mg) were wrapped in polyethylene satchel and attached at selected spots of the cyclotron target-station parts situated in target vault. The locations of the cobalt activation pellets on the cyclotron parts are depicted in Figure 1.

Figure 1: Highlighting the locations of neutron fluence measurement using Cobalt activation pellets: Faraday cup (FC), Switching magnet (SM), Quadrupole lenses (QP), Beam diagnostic ports (BD), Shuttle transfer duct (STD), Shuttle distribution box (SDB) as well as the I-123 (T2.1) and SPECT (TR2.2 and TR2.3) isotope production target stations. The letters in the “( )” brackets indicate the evaluated neutron fluence category as shown in Figure 3.

The Cobalt pellets were exposed to parasitic neutrons produced during twelve days routine isotope production run. During the entire period the proton current bombarding the targets was monitored in real-time using the Health Physics Watchdog [3].

The Cobalt pellets were exposed to parasitic neutrons produced during twelve days routine radioisotope produc- tion run. During the entire cyclotron operation period the

Cyclotrons and Their Applications 2007, Eighteenth International Conference

119

proton current bombarding the targets was monitored in real-time using the Health Physics Watchdog [3] and the data was stored in a database. The integrated proton current (at Faraday Cup) and irradiation time were estimated to be 15489 μAh and 118 h respectively.

The activation pellets were retrieved after 72 hours cool down time (cyclotron shut down period) and assayed using a high purity germanium (HPGe) detector The activity of the 60Co in the cobalt pellet produced via the thermal neutron capture reaction 59Co(n, γ)60Co was estimated. The thermal neutron fluence rate Φ [cm-2s-1] was evaluated using the formula described elsewhere [4].

Φ = Qη-1σ-1N-1[1-exp(-λti)]-1exp(-λtd)-110-24 (1a)

Where

Q = γ-ray count rate of the irradiated cobalt pellet (s-1) σ = thermal neutron capture cross section for 59Co = 37 barn (1barn = 10-24 cm2) N = number of 59Co atoms in the pellet (equation 1b) λ = decay constant (equation 1c) ti = total irradiation time = 118h (Table 1) td = elapsed time between the end of irradiation and counting begin (cool down time) = 72 h

N = Lpkwa-1 (1b) Where

L = Avogadro’s number = 6.02 × 1023 (atoms/mol) p = fraction of element of in a mixture (for 59Co, p =1) k = isotopic abundance (for 59Co, k =1) w = sample weight (g) (for 59Co pellet, w = 0.447 g) a = atomic weight of sample (for 59Co, a = 59)

λ = 0.693/T1/2 (1c) Where

T1/2 = reaction product half life (for 60Co, T1/2 = 5.6 y)

The average thermal neutron fluence φm at seven categories A, B, C, D, E and F were calculated to be 1.5×104, 3.4×103, 1.8×103, 1.1×103 , 8.1×102 and 6.0×102

[cm-2s-1/μAh] respectively.

Building Material of Cyclotron Components We undertook a thorough check up of all cyclotron components located in the target vault (Figure 1) and found following materials as most important building materials: (a) Copper [5], (b) Aluminium-type 5083 [6], (c) stainless steel-types 304, 316, [7], (d) Brass-types 83600, 86300 [8]. The elemental compositions (percent) of the building materials are summarised in Table 1.

Activity Calculations of Cyclotron Components Important nucleonic properties major cyclotron building materials used in this report are presented in Table 3. The activities (A [s-1]) generated in the cyclotron components via the thermal neutron capture is given as:

A = Σn σnNn(1-exp(-λnti))exp(-λntd)Φ (2a)

Table 1: Elemental compositions (percent) of the main cyclotron building materials considered in this work.

Mat. Copper Pipe

Brass (83600)

Brass (86300)

Steel (304)

Steel (316)

Alu (5083)

Al --- --- 7.5 --- --- 93 C --- --- --- 0.08 0.08 --- Cr --- --- --- 20 18 0.25 Cu 99.9 84 66 0.35 0.35 0.1 Fe --- 0.3 4 69 70 0.4 Mg --- --- --- --- --- 4.9 Mn --- --- 5 2 2 1 Mo --- --- --- 0.35 3 --- N --- --- --- 0.1 0.1 --- Ni --- 1 1 10.5 14 --- P 0.1 0.05 --- 0.45 0.45 ---

Pb --- 6 0.2 --- --- --- S --- 0.08 --- 0.03 0.03 ---

Sb --- 0.25 --- --- --- --- Si --- 0.005 --- 0.75 0.75 --- Sn --- 6 0.2 --- --- Ti --- --- --- --- --- 0.15 Zn --- 6 28 --- --- 0.25

Where σn = thermal neutron capture cross section for nth element in the cyclotron part of interest Nn = number of the atoms of the nth element (equation 1b) λn = decay constant of the nth reaction (thermal neutron capture) product (equation 1c) ti = irradiation time [h] td = cool down time [h] Φ = thermal neutron fluence rate (equation 1a)

Φ = φmI (2b) Where

φm = normalised thermal neutron fluence rate of mth category and I = integrated Faraday Cup current [μAh]

A Real Life Example We have calculated the induced radioactivity in a piece of copper pipe, removed from the beam diagnostic port (BD) of target station T2.1 (Figure 1) after a continuous cyclotron operation (isotope production) time (ti) of 28 days, the integrated (Faraday Cup) current (I) was estimated to be 38800 μAh. The copper pipe (w = 500g) was taken out from the target station 2.2 (Figure 1) after a cool down time (td) of 24 hours. The normalised thermal neutron fluence (φm) for category A was estimated as 3.4×103 [cm-2s-1/μAh]. By substituting the values of φm and I in equation 2b, the thermal neutron fluence rate at the location of copper pipe was calculated as:

ΦCu-Pipe= 1.32×108 [cm-2s-1] (3)

By using the list of cyclotron building materials (Table 1), the isotopic abundance of nuclide species (Table 2) in the material of interest and the formula (equation 1b) we estimated the number of 63Cu (N63Cu), 65Cu (N65Cu ) and

Cyclotrons and Their Applications 2007, Eighteenth International Conference

120

31P (N31P) in the 500g copper tube to be 3.27×1021, 1.46×1021 and 9.72×1018 respectively. Furthermore, by substituting the numerical values of ΦCu-Pipe, ti, td, σ (Table 2) and the number of N63Cu (3b), N65Cu (3c) N31P atoms in equation 2a, the activities of the thermal neutron reaction products A64Cu , A68Cu and A31P were calculated as 10.8×108, 0.0 and 10.2×103 Bq respectively. By using the gamma dose conversion factor Γ [9] of 64Cu to be 3.6×10-5[mSvh-1MBq-1m-1] (Table 2) and the activity of 64Cu (10.8×108 Bq), the gamma dose equivalent rate at 1m from the copper pipe was calculated to be 39µSv.h-1. Evidently there was no contribution of 32P (a beta emitter) and 68Cu (T1/2 = 5.1 min), which has completely decayed to nil during the 48 hours cool down period. The gamma dose equivalent rate at 1m from the centre of the copper pipe was estimated with a radiation survey meter was found to be 48µSv.h-1.

Table 2: Important nucleonic properties of the main cyclotron building materials including nuclide species, abundance (a), thermal neutron capture cross sections (σ), daughter product of the (n, γ) reaction, half life and the gamma dose conversion factor Γ [mSv.h-1MBq-1.m-1].

Nuclide Species

a [%]

σ [b]

React. Prod

Half Life

Γ (Dose Const)

27Al 100 0.232 28Al 2.25m 2.4 ×10-4

13C 1.1 0.001 14C 5730y 0 50Cr 4.3 15.9 51Cr 27.7d 6.3 ×10-4

54Cr 2.4 0.36 55Cr 3.5m 0 63Cu 69.2 4.5 64Cu 12.7h 3.6 ×10-5

65Cu 30.8 2.17 66Cu 5.1m 0 54Fe 5.9 2.3 55Fe 2.73y 0 58Fe 0.28 1.2 59Fe 44.5d 1.8 ×10-4

26Mg 11 0.036 27Mg 9.5m 1.4 ×10-4

55Mn 100 13.3 56Mn 2.58h 2.5 ×10-4

92Mo 14.84 0.019 93Mo 6.9h 8.0 ×10-5

98Mo 24.13 0.14 99Mo 2.75d 3.0 ×10-5

100Mo 9.63 0.195 101Mo 14.6m 2.4 ×10-4

58Ni 68.27 4.6 59Ni 7.6×104 y 0 62Ni 3.59 14.5 63Ni 100y 0 64Ni 0.91 1.58 65Ni 2.5h 8.0 ×10-5

31P 100 0.18 32P 14.3d 0 204Pb 1.4 0.66 205Pb 1.5×107 y 6.8 ×10-5

206Pb 52.4 0.49 207Pb 3.3h 0 34S 4.21 0.29 35S 87.2d 0 36S 0.02 0.15 37S 5.05m 0

121Sb 57.4 5.9 122Sb 2.7d 8.2 ×10-5

123Sb 42.6 4.15 124Sb 60.3d 2.9 ×10-4

30Si 3.1 0.107 31Si 2.62h 1.3 ×10-7

112Sn 0.97 0.98 113Sn 115d 4.8×10-5

120Sn 32.59 0.14 121Sn 27h 0 122Sn 4.63 0.18 123Sn 129d 1.0×10-6

124Sn 5.79 0.13 125Sn 9.6d 4.73×10-5

50Ti 5.4 0.177 51Ti 5.8m 7.1×10-5

64Zn 48.6 0.76 65Zn 265d 8.9×10-5

68Zn 18.8 1.0 69Zn 56m 1.2×10-9

70Zn 0.6 0.083 71Zn 2.4m 0

By applying the above method we have estimated the activities in 1kg aluminium (Type 5083) chunks induced

by thermal neutrons during 1 year period (integrated Tgt.current 3.7×105 μAh) presented in Table 3.

Table 3: Predicted thermal neutron induced activation [Bq] in 1 kg aluminium chunks (Test Tag) for area categories A, B, C, D, E and F.

Cr-51 Zn-65 Cu-64 Fe-55 Fe-59

A 4267 1159 473 295 29 B 1051 285 116 73 7 C 446 121 49 31 3 D 327 89 36 23 2 E 245 66 27 17 2 F 178 48 20 12 1

SUMMARY AND CONCLUSIONS A simple experimental method for the prediction of neutron induced component activation in high-current radioisotope production cyclotrons using tiny cobalt (59Co) activation pellets is presented. The inner space of the target vault was divided in six categories based on thermal neutron fluence rate level. Using the thermal neutron fluence rate (evaluated from 60Co activities) and integrated target current we were able to predict the induced radioactivity generated in different cyclotron parts, commonly made of Copper, Aluminium, Brass and Steel. The gamma dose rate near the activated cyclotron part was calculated. The predicted induced activity (Bq) and the resulting gamma dose rate [μSv.h-1] generated in a 500g copper pipe and 1kg aluminium chunks (test tag) exposed in the target vault was confirmed by measuring the gamma dose rate with a radiation survey instrument. This report ignored the activation products of fast neutron and proton induced reactions like, (n, 2n), (n, α), (p, n) and (p, 2n), prevalent in close proximity of targets, like collimator, Haver-foil window and the target station itself.

REFERENCES [1] T. Ishikawa et al. Thermalization of accelerator-

produced neutrons in a concrete room, Health. Phys. 60 (1991)209.

[2] B. Mukherjee et al. (in this Proceeding). [3] B. Mukherjee et al. In Proceedings of the 13th

International Conference on Cyclotrons & Their Applications, Vancouver, Canada, 1992.

[4] W. S. Lyon, Guide to Activation Analysis, D. Van Nostrand Co. Princeton, 1964

[5] Standards Australia AS 1432-1990, Standard Associa-tion. Australia, North Sydney

[6] Standards Australia AS 2848-1986, Standard Associa-tion Australia, North Sydney

[7] Standards Australia AS 1432-1990, Standard Associa-tion Australia, North Sydney

[8] Standards Australia AS 2738-1984, Standard Association Australia, North Sydney

[9] B. Shielen, Health Physics and Radiological Health Handbook, Scinta Inc. Silver Spring, 1992

Cyclotrons and Their Applications 2007, Eighteenth International Conference

121